240 PROFESSOR W. A. TILDEN ON THE SPECIFIC HEATS OF METALS, 



The specific heat of a metal is affected by a variety of other circumstances, of 

 which its mechanical condition is the most important. REGNAULT found the specific 

 heat of hammered copper '0935 and that of annealed copper '0952. According to 

 my results porous unfused metals all gave slightly lower values for the specific heat 

 than the same after fusion. 



As explained at the outset, all the determinations described up to this point were 

 made in the steam calorimeter, and the results represent the average specific heats of 

 the metals between the temperature of air, say 15, and that of steam 100. On 

 applying the rule of DULONG and PETIT it has been shown that, notwithstanding the 

 high degree of purity of the metals nickel and cobalt, the extreme care taken in deter- 

 mining their specific heats under the same conditions, and the close approximation of 

 the physical properties of these two metals to each other, the numbers representing 

 the atomic heats within the range of temperature of 15 to 100 differ appreciably. 



The specific heat represents the relative amount of energy consumed in giving 

 to the molecules of the solid the vibratory motion corresponding to temperature 

 and in separating the molecules from one another, and so causing expansion. 

 I am indebted to Mr. A. E. TUTTON, F.R.S., for determinations of the coefficients 

 of expansion of the two pure metals, and his results have been published in the 

 'Proceedings of the Royal Society' (vol. 65, pp. 161 and 306). He finds the 



. for nickel 10~ 8 (1248 + 1'48) 



coefficient 01 linear expansion at t i. ,A_*) r.\> a difference of about 



for cobalt 10~ 8 (1208 + 1'28<) 



3 '2 per cent, at 0. His determinations also indicate that this difference increases 

 with rise of temperature, and amounts to 4*3 per cent, at 100. These differences of 

 expansibility correspond very closely with the differences in the specific and atomic 

 heats of the two metals. 



I have also to thank Mr. THOMAS TURNER, Assoc. R.S.M., F.T.C., for observations 

 with his sclerometer,* on the relative hardness of the two metals. He reports that 

 the cobalt is harder than the nickel, the relative hardness being on his scale about 

 Co = 17 and Ni = 12 or 13. The cobalt is harder than soft wrought iron, while the 

 nickel is softer than wrought iron. The difference is therefore very distinct. 



These differences of expansibility and of specific heat are the result of observations 

 at and above atmospheric temperatures. As the specific heat of solids increases with 

 rise of temperature and diminishes with fall of temperature, it was thought desirable 

 to make some estimations of specific heat at successively lower temperatures in order 

 to find out if the difference between the two metals was maintained. At absolute 

 zero it is probable that they would exhibit the same capacity for heat, and if 

 temperatures low enough could be employed an estimate could be made of the 

 absolute atomic heat of the solid metals. 



A series of calorimetric experiments has, therefore, been made at the tempera- 



* ' Journ. Chem. Soc.,' 1885, vol. 47, p. 904; and 1887, vol. 51, p. 145. 



